News blog

ADHD gets the attention of geneticists


True to its name, attention-deficit hyperactivity disorder (ADHD) has had a tough time garnering much notice from behavioural geneticists. The condition is highly heritable, but genetic association studies of ADHD, which affects about 2% of children, have mostly come up blank – until now.

A team led by Nigel Williams and Anita Thapar at Cardiff University School of Medicine in the UK have just linked the condition to large missing and extra chunks of chromosomal DNA known copy number variants (CNVs). “It gives us the first genetic link to ADHD,” Thapar said at a press briefing this morning.

Her team scanned the genomes of 410 children from Wales and northern England with ADHD and 1156 adults who were part of a large cohort study, looking for chunks of DNA 500,000 letters or longer that were either deleted or duplicated. They also limited their scan to rare CNVs, found in less than 1% of the general population.

On average, the children with ADHD were about twice as likely to possess such a mutation, compared to controls. Those children with an IQ below 70 were even likelier to have a swath of their genomes missing or duplicated.

CNVs have also been linked to autism and schizophrenia , and the results from the new study turn up some of the same chromosomal regions as those studies, such as a patch of chromosome 16 that is prone to duplication. Williams and Thapar’s team says this overlap could suggest a shared biological basis for autism and ADHD.

For the most part, however, it’s difficult to determine how the CNVs Thapar and Williams’ team identified are related to ADHD. Most of the insertions and deletions contain a number of genes, some implicated in brain function, some not.

Stephen Scherer, a medical geneticist at the Hospital for Sick Children in Toronto, Canada, expects that some of the CNVs will almost certainly play a causative role in an individual’s ADHD. An easy way to tell, he says, is to identify so-called de novo, mutations found in children with ADHD, but not their parents. In an analysis limited to a subset of their participants, Thapar and Williams identified four such mutations, including one on chromosome 16.

Like most genetic associations, the clinical applications of new study are years off and not obvious. “Williams and colleagues results are exciting, but how these findings will be clinically translated is still speculative,” writes J. Peter Burbach, in an editorial that accompanies the paper, which both appear in the journal Lancet. At a London press briefing, Thapar batted down question after question on the real-world application of her team’s work – but for one.

She hopes her team’s study will dispel the popular perception that ADHD results from poor parenting. “Finding this direct genetic link to ADHD should help clear this misunderstanding and address this issue of stigma,” she said.

Image: photo by tonx via Flickr under Creative Commons.


  1. Report this comment

    Josephine Elia, MD said:

    On July 23, 2009, the first study indicating a causative role for rare structural variants in ADHD was published in Molecular Psychiatry by our group (Elia, Gai, et al., Molecular Psychiatry; 15:637-646; The results from our study as well as similar work by others has prompted the scientific community to switch paths from looking for common variants in these complex neuropsychiatric disorders, such as ADHD, to adopting research models that also investigate rare variants.

    Thus we are encouraged to begin seeing publications of studies investigating rare variants in ADHD such as the one by Williams et al. in The Lancet. From our observations, the Williams study comprises two major lines of investigation and conclusion. The first is that large structural variants observed in individuals with ADHD are preferentially observed in patients with autism and schizophrenia. As with our published study, the relatively modest cohort size and high degree of genetic heterogeneity has as yet prevented the implication of individual ADHD risk genes with statistical significance. However, it is notable that while a few of the genes with large CNVs observed exclusively in the ADHD cohorts were the same as in our study, including AUTS2 and PARK2, most genes were different, consistent with the hypothesis that a substantial number of loci contribute to ADHD risk. It would be most interesting to see the degree of overlap in ADHD-enriched genes between the studies when including smaller CNVs from the Williams data. In this regard, it is unclear to us why the authors excluded smaller CNVs from consideration. The authors indicate that this was performed due to the propensity for larger CNVs to be associated with disease risk in other neuropsychiatric disorders, but this fact is likely due to ascertainment bias and that larger CNVs generally overlap a proportionally larger number of genes. We suggest that a gene-based analysis rather than reliance on a size cutoff may be more appropriate in this context. The paper, nonetheless, lacks any systematic analysis of the genes implicated by these CNVs. Regardless, the indication of enrichment for known autism and schizophrenia genes/regions in the ADHD CNV population is consistent with our prior findings.

    The second line of investigation concludes that there is a higher frequency of CNVs associated with ADHD. However, we believe that the data presented does not support this conclusion. The manuscript reports 5.6-fold enrichment of large CNVs in the ADHD subjects with IQ70. However, the mean IQ of those ADHD patients with IQs >70 is only 89, whereas an expected mean IQ for a proper healthy control should be ~100 (the study does not report control IQs). In contrast, the mean IQ in our cohort was 104, and we did not observe an increase in deletion or duplication frequency in our ADHD cases. A more appropriate approach would be to closely match cases and controls for IQ, or to at least recalculate significance for a subcohort of cases where the IQ mean approximates 100. In the absence of this correction, establishing a direct association between ADHD and large CNV prevalence is difficult due to the possibility of intellectual ability as a confounding variable. Stated more succinctly, the significant increase in CNVs observed in the ADHD individuals “without intellectual disability” may be attributed solely to a higher rate of CNVs in individuals with IQs in the low normal range.

    Notably, our study established an enrichment of ADHD-associated CNVs in pathways that are important in ADHD, such as learning, behavior, synaptic transmission and central nervous system development. However, we have also uncovered the same CNVs reported in the study by Williams et al in a cohort of children with developmental delay and/or mental retardation but without ADHD or autism, suggesting that these CNVs may be more relevant for cognition rather than ADHD.

    Finally, the authors claim that these findings refute the hypothesis that ADHD is purely a social construct. While we consider this message to be of utmost importance for the ADHD community, we suggest that the Williams study contributes to a vast amount of prior evidence derived from brain imaging, pharmacological, genomic, and family and twin genetic studies that clearly indicate a strong genetic contribution to ADHD risk, and that this prior work should be appropriately acknowledged. We look forward to seeing studies that implement rigorous scientific methodology to further elucidate the molecular basis of ADHD.

    Josephine Elia, MD, Xiaowu Gai, Ph.D.,

    Hakon Hakonarson, MD, Ph.D., Peter White, Ph.D.

  2. Report this comment

    Lis Jessie said:

    Hi! I was wondering if the genetics of ADHD in high IQ individuals has been studied at all? It would be interesting to see how the CNVs differed from control and low groups. I am a woman who inherited a severe form of ADHD from my father along with his ridiculous IQ. I’d be grateful if you could point me to any specifically related research. It might provide me better insight I could share with my neurologists (if they’ll listen) Thanks for your time!

Comments are closed.